On 20 December 2017, the European Commission published the long-awaited draft of Annex 1 “Manufacture of Sterile Medicinal Products” of the EU GMP Guide. A public consultation period will run until 20 March 2018.
What’s new in the draft? Last week in LOGFILE 07/2018 you already read the first part of the detailed analysis that was carried out by our authors Ruven Brandes and Fritz Röder. This second part highlights what problems remain and which are entirely new. It provides a comparison with the FDA Aseptic Guide and concluding thoughts on the draft.
Despite the pledge to eliminate ambiguities and inaccuracies from the previous version, in the authors’ opinion some problems remain.
The requirement for linking of the bioburden limits with the filter efficiency of the sterile filter (section 10.2), taken literally, would result in limits which an inspector would not accept.
The statement: “The final stage of the airlock should, in the at-rest state, be the same grade as the area into which it leads” (section 5.9) with regard to the final room grade in an airlock remains unchanged even though it has presented a significant problem for years.
The differences between the conditions for grade A and the air supply for grade A remain unclear. Unfortunately, the definition of grade A air in the glossary does not fully resolve the ambiguity.
The requirement that a sample is taken from the worst case sample point in a water system each time the water is used for production (section 7.15) is extremely problematic and hopefully unintentionally misleading. In light of the fact that some systems may be used 100 or even 1000 times a day, that would be a lot of water analyses!
When producing pure steam, “purified water” at a minimum is required as feed water. In the opinion of the authors, a reverse osmosis permeate with endotoxin monitoring/reduction would suffice for production. This would have the benefit that this water system would not be subject to any pharmacopoeia requirements.
Although is it is not really a crucial issue, the definition of single use systems is not very helpful and could be improved. Here, it would be useful to actively seek clarification with the manufacturers.
There are a few small typographical and language problems, which will hopefully be solved during the consultation period. Some have implications for the meaning of the text, while others are unnecessary. For example, actions can be termed as “preventative” unless the guidelines are expressly talking about CAPA. In this case, they are “preventive”.
In the European Annex 1, in contrast to the FDA Aseptic Guide, there are requirements for the crimping process as well as differentiation between “aseptic” or “clean processing”. For the latter, a more precise definition of the “grade A air supply” which, according to Annex 1, should be used for protection during the process, would be helpful. Unfortunately, this more precise definition has failed to materialise. A widely recognised standard for this exists in the industry (use of filtered air in accordance with grade A without consideration of the microbiology), but this has not found its way into Annex 1 to date.
There are significant differences between Annex 1 and the Aseptic Guide when it comes to sterilisation. In the US document, there are no references to terminal sterilisation. This is included in the new EU document. Limiting the requirement for sterilisation with pure steam primarily only to terminal sterilisation and also allowing sterilisation with ethylene oxide or other methods for so-called ready-to-use materials is implemented in the draft, which should be seen as positive.
There is also a difference in quality oversight. In Europe, there is no requirement for quality assur-ance for aseptic processes to (physically) be on-site. In contrast, the Aseptic Guide requires QA oversight, media fill in particular is specified here. This alignment has not been carried out. The mini reference to supervision by a quality unit during gowning, in the context of personnel qualification, does not have the desired scope.
The revised Annex 1 is, as promised, a comprehensive revision of the previous version. The focus is on the application of QRM and PQS in sterile manufacturing. The designation and description of the use of modern barrier technology as the state of the art for aseptic manufacturing has been very well implemented in the draft.
There is some new, excellent information regarding a wide range of technologies which were previously not sufficiently covered. Some ambiguities and inaccuracies from the 2008 version have been rectified.
DIN ISO 14644-1, which is used by both the European and the American guide for the classification of cleanrooms, is a central document in the draft. The fact that the limit of 5 µm particles has been removed from the ISO 5 class (ISO 4.8) should be viewed positively. Discrepancies for 0.5 and 5 µm particles run essentially parallel, so the limit for 5 µm particles can be foregone. The change with regard to 5 µm particles in routine operation closes a regulatory gap, but still allows the manufacturers to continue measuring this particle fraction. Which is now consistent since an increased number of these particles indicates a problem.
An opportunity for improvement was missed on the topic of sterile filtration. Integrity testing after sterilisation immediately before filling could have been left out. The data from filter verification and the integrity testing after filling provide sufficient assurance. As a result, dispensing with mandatory integrity testing after sterilisation (before use) would have been desirable. Dispensing with this requirement would reduce the complexity in the aseptic set-up and in the design of facilities.
Another theme in Annex 1 is verification of the integrity of containers with sterile medicinal products. Currently, Annex 1 only requires 100% verification for containers which are closed by fusion (glass ampoules and BFS containers). It is surprising that more openness towards suitable controls for certain types of packaging and dosage forms has emerged here.
This draft has taken a significant step in the right direction.
Of course, as for all new publications, and particularly drafts, there are points of contention. The public consultation can and should address many of the points and thus provide potential improvements.
Wirtschaftsgenossenschaft deutscher Tierärzte eG (WDT), Garbsen, Germany
Merck KGaA, Germany